Future Possibilities

Turn of the Century Weight Loss Texts & Current Weight Loss Verticals

As noted on the Analysis page of this site, this study compares the text of books with that of a website, the latter consisting mostly of headlines, so the type of texts being compared is significantly different. Time is also an issue in this study, as the weight loss verticals are both from one day in the year 2019, while the turn of the century corpus spans a comparatively huge time frame of thirty years. Future studies might achieve of clearer one-to-one comparison by constructing a current-day corpus consisting of full articles or even diet and exercise books comparative to the turn of the century texts. The difficulty would be to control for bias when selecting said articles/books, as the cultural corpus has grown exponentially since the 1920s, as have the various tool and activities associated with weight loss.

This brings us to another issue, which is the comparatively small size of the turn of the century weight loss corpus: eight texts. While this creates a higher word count than the weight loss verticals, it represents fewer voices, since the verticals each feature dozens of articles by a variety of authors. Yet these were the only texts found using this study's methodology for searching on HathiTrust. Further research might investigate the variety and quantity of weight loss texts printed during the time period in question, as well as consider either alternate modes of searching HathiTrust or alternate sources of texts.

HTRC Analytics & Voyant Tools

These tools were both chosen for the ease with which they could process their respective corpuses, which are quite different from each other in form: one being static text digitized with OCR, the other being changeable, scrolling webpages. In terms of ease, both tools performed beautifully. However, though it would have been more difficult from a technical perspective, a more detailed close comparison could have been made by processing both corpuses through the same exact tools.

This became especially clear when examining gender through the lenses of Voyant's Trends tool and HTRC Analytic's InPho Topic Model Explorer (see the Analysis page of this site for more on this examination). While the InPho Topic Model Explorer is the more elegant tool, it ended up revealing less about the texts in the corpus than the Trends tool - or, for that matter, than either Voyant's Cirrus and Summary tools or HTRC Analytics' Token Count and Tag Cloud Creator. Voyant has its share of iffy tools as well, but its wide variety still provides many more analysis options than HTRC Analytics does. It's possible that some Voyant tools that proved unnecessary for this particular study - such as Bubblelines, Collocate, and Knots - could have been made useful if both corpuses under consideration had been processable through these tools. Future research might consider finding a way to process both turn of the century texts and current materials via Voyant Tools or a similar multi-tool package.

Introduction | Background | Methodology | Results | Analysis | Future Possibilities | References